The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from URLs and tool definitions, creating a vulnerability surface where malicious instructions in that data could influence the agent.\n- Ingestion points: External content is ingested from URLs, rendered observations, and discovery files as described in SKILL.md.\n- Boundary markers: The skill lacks explicit markers or instructions to ignore instructions embedded in the external data.\n- Capability inventory: The instructions involve file access and tool execution across the defined workflow steps.\n- Sanitization: No sanitization or validation logic is present to filter malicious content from the external inputs.

seo-agentic